After a general introduction to machine learning and TinyML, we will dive into a concrete use case in the domain of audio event detection. 

We have recordings from a fitness center that uses a bell sound mixed into an audio stream to indicate a 45-second interval and to notify the users to move on to the next machine in a loop of 16 workout machines. 

We need a small microcontroller to synchronize the 45-second bell sound with a light animation that will help people training with headphones to know when to move to the next machine. 

What to expect:
During the workshop, we will ensure that we all get through the various steps from data gathering, audio labeling, and preparation to model selection, training, and deployment. 

We will work on the ESP-EYE module - a small, powerful dual-core processor with a built-in camera and microphone. 

During the design and implementation process, we go through a couple of different approaches to do the data labeling, training frameworks, and deployment options. 

If you have a relevant use case, you are more than welcome to work on that, alone or in a group.

We will be a couple of instructors from IDA embedded with domain knowledge from audio, vision, machine learning, and advanced embedded programming.

The workshop will run in parallel in Copenhagen  

Link to githup repo:
Will follow soon

Tools:
Audacity,  Tensor flow lite, Python, ESP IDF 

All workshops and dates
This is the second of 3 workshops, that is also organised and streamed in Copenhagen. You can sign up for one of all three workshops here:

November 20 - Introduction to TinyML
Workshop 1 - IDA København
Workshop 1 - AU Herning

December 4 - TinyML Data preparation, training and deployment
Workshop 2 - IDA København
Workshop 2 - AU Herning

December 18 - TinyML Production model development
Workshop 3 - IDA København
Workshop 3 - AU Herning